Thread: Air speeds in and around duct openings and blades (NB Invisible dust discussed)

Originally Posted by Chris Parks

A bloke called David Vizard did a lot of flow bench testing of different available types of ram tubes with diagram and flow results back in the seventies for a carburettor article. Sadly my copy went to the tip years ago. I guess some clay or putty, a good controllable air souce and a manometer would yeild results. A flow bench would be better of course.

Sounds like a possible student project to me

Originally Posted by Chris Parks

A bloke called David Vizard did a lot of flow bench testing of different available types of ram tubes with diagram and flow results back in the seventies for a carburettor article. Sadly my copy went to the tip years ago. I guess some clay or putty, a good controllable air souce and a manometer would yeild results. A flow bench would be better of course.

I still have his book in a box somewhere along with Smokey Yunick's Power Secrets, who was another pioneer... Dave Vizard also did some interesting work on anti reversion headers (placed cones inside the exhaust headers to allow the gas to pass one way, but got trapped when trying to re-enter the cylinder head.

Back in the Holley days (pre fuel injection) we used to cut the choke housing off and use an araldite type substance, to build up that area to make a pseudo velocity stack.

Last week Interwood and I investigated the air flow in the vicinity of a couple of different hoods that are commonly found in hobby level DC accessory kits.

The one on the left is symmetrical in shape with a flattish 310 x 310 mm extension
The one on the right is asymmetrical with the longer lip being 87 mm and the shorter one being 25 mm

Both hoods have 90 mm ID inlets which makes them difficult to mount on 104 mm ID ducting.
In the end we padded out the outside of the hood inlets and jammed them into the 100 mm ducting.
This is when we observed something very interesting and that was the hood kept jumping off the ducting. Ordinarily we would have expected the hoods to be sucked onto the ducting inlet but no they kept jumping off.

Here is a video showing this effect and a couple of other things.
BobL Dust Hood Measurements 22 Aug 2012 - YouTube

This is a great example of the Bernoulli effect in action where the faster moving airstream across the front of the hood and into the inlet creates a lower pressure than the air behind it .

The implication for any DC is that this uses up valuable suck (or generates resistance) that would otherwise go into moving air. The other problem with using these 90 mm ID tapered inlet hoods and also the blast gates is that they represent a further constriction to flow and also generate turbulence in 100 mm ID PVC pipe.

Anyway we eventually managed to get the hoods to stay on the ducting and we collected a fair bit of data. Rather than drown everyone in all the data we propose to just show a few comparisons

This is the air speed distribution across the bigger hood.


The main thing to observe here is the relatively higher air speed at 100 to 150 mm from the front of the inlet. This is obviously because the inlet cannot draw air as easily from behind the hood.

The next graph shows some air speeds for the smaller asymmetric hood. To keep things simple we only show the air speeds directly out from the inlet and at 45º across the longer and shorter sides of the hood.


In this case the difference in air speeds 45º out for the shorter and longer sides is very clear with the longer (more protected) side having the higher airspeed – just like the large hood.

IN the following graph we compare the airspeeds and how they vary with distance directly from the front of ;
- a plain 100 mm ducting inlet
- the small asymmetric hood
- the larger hood.



This shows that between 100 mm and 200 mm the bigger hood generates higher air speeds than the smaller hood which in turn has a higher sir speed than the plain ducting. I don’t know why this is reverse at 50 mm.

It should be borne in mind that, while the air speeds are higher for the hoods than the plain ducting the total air flow through the hood inlets is almost certainly lower as the plain 100 mm duct has an inlet diam of 104 mm while the hood inlet is 90 mm in diameter. The plain ducting is drawing a considerable amount of air from behind the inlet. This was clearly demonstrated by placing the anemometer a mm or so behind the inlet where air speeds up to 10 m/s were measured.

What is the bottom line here. One way of putting it is, want to collect the most dust? If so then hood may help to generate higher air speeds in the vicinity of the dust making space, but to vent a shed, remove the hood and more air will flow and empty the shed quicker that using ducting with hoods.

Great post Bob

What we need to know (I guess) is what is the optimal shape hood to use for this purpose - a funnel shape? If so, what is the best angle for the funnel?

A smoke flow would be interesting to see.

We'll be into CFD next. It is about time someone has finally taken the time to do some practical demos. Good one Bob.

Cheers guys. Before I forget I should thank Interwood for taking the vid an also providing the hoods.

Originally Posted by Groggy

What we need to know (I guess) is what is the optimal shape hood to use for this purpose - a funnel shape? If so, what is the best angle for the funnel?

That would indeed be a worthwhile experiment. Maybe something like those barn door stage light arrangements attached to the sides of an inlet might reveal something of interest.

Given that all hoods generate some resistance, if for what ever reason you want to grab the most air (ie eg shed vent) then the most efficient shape is almost certainly the bell mouth shaped inlet. The same goes if you don't mind big chips escaping but want to grab invisibles.

If you have a very localized dust producing process (eg drilling) then something that generates the highest local air speed (maybe even like a vacuum cleaner nozzle) could still be more effective as an all round performer (chips and invisibles). Even though I suspect it will be quite local I'd like to see some air speed mapping while drilling.

Some form of smoke flow is on the agenda.

Had a spare half hour this evening to do a bit more testing - this time the table saw.

Here is the item in question - 12" contractor TS with attached router table
Custom hopper with 6" DC inlet in base of saw.
Back almost closed off completely
Custom aluminium zero clearance throat plate.
Adjustable height hood made from a dust picker and some PC sheet.

Hood has 4" inlet above table
DC is pulling about 1200 ccm shared between the 6" and 4" inlet.

I borrowed Doug's idea for the photos of using paper streamers to show air flow
I placed a streamer at the front, back and side of the hood all facing outwards like this (nothing is on at this stage)

The hood is ~25 mm above the table top

All speeds shows are the highest measured in the 25 mm gap.
Near the table the speeds are about 20% less than the max speed.

Here is the TS on only (no DC)

Air flows into hood at back and sides and OUT at the front at speeds shown.
I think the blade is acting like a scoop - it grabs air at the back and as it goes around and spills it at the front.


Here is the DC on only (Saw off)

Air flows more or less evenly in around back - front and side of hood.
The gap between the hood and Table is not even so the speeds are probably just gap width dependent.

Now - here are the TS and DC both on

ALL speeds are INTO the hood
Note slight decrease in speed at the front and increase in sides and front.
BUT the speeds are not additive !
This demonstrates that if the DC extraction volume is large enough it can overpower the forward air spill produced by the blade - this is very significant

I also measured air speeds at the gaps shown

A = 3.5 - 3.7 m/s with just the DC on and saw off and it dropped slightly when both were on t0 3.4 - 3.5 m/s
B = 3.5 m/s and did not change whether the saw was on or not.
All directions were into the cabinet.


Here is the back of the saw as you can see it's mostly close as I rarely do mitre type cuts and when I do I just take the back off.
The air speed here with DC running and TS on or off was 3.2 m/s into the cabinet
Note MDF dust from doing experiments with no DC running (I should get rid of that)



So what does it all mean.
It demonstrates the need for a good hood that covers the work and can be tucked down nice and low otherwise that air current off the blade will send dust forward toward the operator - we've all seen that
It says nothing about what air currents will be generated when big chunks of stuff start flying off.

Lots more experiments to do, different blades, different heights above work, real live cutting - anything else?

Originally Posted by BobL;1552508...

I borrowed Doug's idea for the photos of using paper streamers to show air flow...

Thanks for incorporating my streamers into the photos, Bob. It makes the pictures more meaningful. The way you combined the flow readings you took with the visual representation of the way the streamers are affected gives us a figure that can be plotted on a graph as well as a visual representation of what that figure looks like in real life. Still pictures wont do it as much justice as a video would but it does add that extra something that makes it tangible more so than just a figure on a meter and it is not always easy or practical to record or post a video.

Just a thought, Bob, can you run some kind of experiment to determine what strength of airflow is required to capture the majority of invisible airborne dust in an open area as opposed to in a duct, and then create a visual of what that airflow looks like on a paper streamer? I know that there are a lot of variables in what I just asked but if you think about it as, say, " if a single ply strip if kleenex tissue 5mm wide and full width of the tissue out of the box (ok now everyone knows where i get my paper streamers from) is deflected by xx degrees towards the machine when measured at xx cm from the machine the dust collection is producing enough suction to capture xx percent of the invisible dust".

Another thought I had about invisible dust: If I found a filter that only collects invisible dust, how could I see when it was full?

Doug

I still like the brush sealing idea with a twist. I think that if an entry were made at the top of the guard to allow make up air to be pulled in it would work as well as anything. The brushes are guaranteed to form a good effective barrier to dust escaping at high velocity and if it wanders around a bit under the hood it does not matter. I haven't gone ahead with making one as I am moving to another saw. The biggest issue with over head hoods as I see it is allowing enough air in to match the extraction requirements as all the possible entry area gets blocked off if the hood rides and thus seals on the timber being cut. Looking at the Shark Guard I have never understood how it is as effective as those who use it reckon it is. Maybe I am missing something.

What if the air was drawn from the side across the blade? A good air flow might disturb the propensity for the blade to carry the debris? Just an off the wall thought.

Originally Posted by doug3030

Thanks for incorporating my streamers into the photos, Bob. It makes the pictures more meaningful. The way you combined the flow readings you took with the visual representation of the way the streamers are affected gives us a figure that can be plotted on a graph as well as a visual representation of what that figure looks like in real life. Still pictures wont do it as much justice as a video would but it does add that extra something that makes it tangible more so than just a figure on a meter and it is not always easy or practical to record or post a video.

Cheers Doug. I was going to take some video but all I needed to show was a direction.

Just a thought, Bob, can you run some kind of experiment to determine what strength of airflow is required to capture the majority of invisible airborne dust in an open area as opposed to in a duct, and then create a visual of what that airflow looks like on a paper streamer? I know that there are a lot of variables in what I just asked but if you think about it as, say, " if a single ply strip if kleenex tissue 5mm wide and full width of the tissue out of the box (ok now everyone knows where i get my paper streamers from) is deflected by xx degrees towards the machine when measured at xx cm from the machine the dust collection is producing enough suction to capture xx percent of the invisible dust".

I have never had much luck with using paper streamers as a speed indicator as they only work over a narrow speed range. They just don't move when the air speed is too low and then they flick around like crazy when the air speed is too high. Groggy has suggested using wool strands which I don't have in the shed but SWMBO is sure to have some in her Tardis somewhere.
I'll try to quantify this inn some way.

FWIW any air speed will capture invisible dust - but of course the slower it is the longer it takes and the more exposure the shed operator receives.

Here are a couple of examples.
A 20 x 16 x 9 ft shed has a volume of 2880 cuft
400 cfm ventilation (the best one can hope for with a 1HP DC and 4" ducting) wil ltake 7.2 minutes to perform one room air change. At best (with even cross ventilation) only half the invisible dust will be removed in that pass. The next pass reduces it to 1/4. The next to 1/8th etc. 10 room changes (72 minutes) reduces the dust level to 0.2% of the original amount and is considered a target to aim for.
1200 cfm will only do this 3 times faster (24 minutes).
Of course this assumes good even cross ventilation and no other dust making activities in the meantime.
For typical shed ventilation via a small door or widow multiply all the above by X
This is why "capture at source is considered the go"

Another thought I had about invisible dust: If I found a filter that only collects invisible dust, how could I see when it was full?

It's only visible when it is suspended but like talcum power when it settles it will clump so it can then be seen. How will you know a filter is full is because it will cease passing air. This is when a manometer comes in handy because it's resistance will also increase dramatically

Originally Posted by Chris Parks

I still like the brush sealing idea with a twist. I think that if an entry were made at the top of the guard to allow make up air to be pulled in it would work as well as anything. The brushes are guaranteed to form a good effective barrier to dust escaping at high velocity and if it wanders around a bit under the hood it does not matter. I haven't gone ahead with making one as I am moving to another saw. The biggest issue with over head hoods as I see it is allowing enough air in to match the extraction requirements as all the possible entry area gets blocked off if the hood rides and thus seals on the timber being cut. Looking at the Shark Guard I have never understood how it is as effective as those who use it reckon it is. Maybe I am missing something.

What if the air was drawn from the side across the blade? A good air flow might disturb the propensity for the blade to carry the debris? Just an off the wall thought.

I still like the bristle sealing idea as well - I'm awaiting the outcome of these and other tests before I decide what to do with my bristles. I like the idea of the opening at the top - I might just give that one a go.

I tried some wool as an airflow indicator.
Don't ask me about the colour - its what SWMBO had spare.
Anyway I said you boys would like it.

DC on (TS on or off makes no difference visually)
TSandDCon.jpg

TS on DC off.
TSonandDCoff.jpg
The cross over between in and outflow of the hod can be see seen on the side.
It appears to be level with where the blade passes through the throat plate.

The wool did not appear as sensitive as the tissue paper to indicating air flow direction although some sort of neon coloured wool would make it easier to see..

Cheers
Bob

Originally Posted by BobL

I tried some wool as an airflow indicator.
Don't ask me about the colour - its what SWMBO had spare.
Anyway I said you boys would like it....
Bob

Well it doesnt look like it is actually pink, is it Bob?

Maybe a manly shade of puce, but DEFINITELY NOT pink!!!

Doug

Originally Posted by pjt

This thread has jumped way ahead of where I was upto a few nights ago so I haven't got to the end of the thread but a couple of things come to mind here, a clear example of a bell mouth on an inlet is the air inlet for fuel injection/carb for race car engines, just look at Jack Brabham's cars from back in his day.
Further to this inlet discussion (for a square end) the bell shape tends to form inside the pipe therefore reducing the effective diameter of the pipe so if a bell shape is moved outside the pipe (a bell end) we obtain closer to full airflow inside the pipe as this directs airflow into the pipe with a smooth transition, Bob might be able to find a pic/ref, I tried but


Pete

Bit late know, just came across some pics. Here's a pic of some from a car I restored many years ago.